Ian Lynagh igloo@earth.li
Mon, 24 Sep 2001 16:03:17 +0100

On Mon, Sep 24, 2001 at 07:41:42AM -0700, Simon Peyton-Jones wrote:
> | Just to make sure I am interpreting the report correctly, is 
> | ":" meant to be a valid qconop regardless of the 
> | precedence-level and associativity of the qconop?
> I don't really understand the "regardless of.." part.  But, yes, 
> ":" is a qconop, with precedence and associativity infixr 5.

Ah, I was reading it wrongly then.

Can I suggest it (the abstract syntax) is rewritten in the revised
report along a similar line to how - is handled in expressions? In fact,
I think making explicit all the things which can take a precedence or
associativity would be useful.